Classifications Of Customers And Power Network Cost Allocation Among Direct Trades Between Generation Companies And Large Consumers

Promoting direct trades between large consumers and generation companies can explore the potential of an electricity market. Direct trades between large consumers and generation companies have impacts on power system planning and operation, and brings new challenges and opportunities to power industry. In order to guide the sustainable development of power industry and optimize allocation of resources, it is vital to scientifically design the principle of direct trades. At present, the main problems existing in domestic pilots of direct trades are:the understanding of all parties to the nature of direct trades is not clear, the market access mechanism to direct trades is unreasonable, the independent transmission and distribution price has not yet formed, the implementation of direct trades affects the market fairness and so on.In view of the above issues, in-depth research is performed in the following aspects, with respect to the situation and characteristics of direct trades between large consumers and generation companies at home and abroad.(1) In view of the problems existing in direct trades, some suggestions are put forworded in market access, transmission and distribution price, trade category, quantity of trading electricity and trading mode, in order to provide the reference for further development of direct trades. And on this basis, four developmental stages of direct trades and the characteristics of each stage are summarized.(2) For the purpose of determining admissible large consumers in each developmental stage, a consumer priority classification model based on Self-Organizing Mapping (SOM) and K-means clustering is proposed. Firstly, evaluation index system of large consumer priority is constructed from characteristics of power consumption, environmental protection conditions, credibility and load importance. Then, a novel clustering method which combines self-organizing neural network and K-means is employed to classify the consumers. At last, an improved Technique for order preference by similarity to ideal solution (TOPSIS) approach is applied to evaluate the priority of every group of consumers.(3) Annual permitted revenue of a power enterprise must be determined first in order to form the transmission and distribution tariff mechanism. It is vital to form stable annual permitted revenue to forcast power network investment demand. The power network investment demand could present significantly different characteristics at different development stages, and how to accurately forecast this demand is a very important issue for a power grid company so as to develop the optimal investment strategy, optimal planning scheme for the power network concerned, as well as the optimal operation strategy. Given this background, according to the relationship between the power network investment demand and development stages, an adaptive bell curve model with an environmental factor is first presented to describe the investment demand curve. Then, the effectiveness of this model is demonstrated by using the actual historical investment data of Japanese power industry in the period from1955to2010. Finally, the investment demand of a provincial power grid in east China is investigated by employing the proposed model.(4) How to allocate power network fixed cost is an important but not yet well-solved issue related to direct trades between large consumers and generation companies. Given this background, this issue is systematically analyzed considering various attributes of direct trades. Specifically, an evaluation system for allocating power network fixed cost is structured first from three aspects: the user side, generator side and the impacts on the power system. Then, a cooperative game model in the framework of data envelopment analysis is presented to address this issue. In this model, weight constraints are introduced by employing the interval analytic hierarchy process (IAHP) to reflect the preference of decision-makers, and a power network fixed cost allocation model considering the index weight constraints is next attained.Finally, several conclusions are obtained based on the research outcomes, and directions for future research indicated.